P520_BBA5_pf

csimoncsimon Acadiana Icrontian
edited March 2004 in Folding@Home
I'm trying to make an observation here ...I'd like to figure out how many mhz/ghz of amd (32bit) processor will fold this protein at appoximately 6:34 - 6:40. At 2.4g the amd must do better that this I'm sure. I'm thinking around the 2g amd range will be equal to the 2.4g p4 in folding but I may be wrong.

This is what I have so far:

amd 2000+@1.75 = 7:12
p4 2.4g = 6:37

Comments

  • shwaipshwaip bluffin' with my muffin Icrontian
    edited March 2004
    6:35 = 395s

    7:12 = 432s

    395*f = 432 * 1.75

    f=1.913ghz
  • mmonninmmonnin Centreville, VA
    edited March 2004
    AMD 2.5 = 5:20 a frame
    AMD 2.4 = 5:24 a frame
    AMD 2.3 = 5:34 a frame
    AMD 2.1 = 6:15 a frame
    AMD 1.7 = 7:48 a frame


    Thats what EM3 says. So I could say about 1.8-1.9Ghz
  • csimoncsimon Acadiana Icrontian
    edited March 2004
    so roughly p4 2.4gig is comparable to amd 1.85gig fps or thereabout?
  • shwaipshwaip bluffin' with my muffin Icrontian
    edited March 2004
    on that protien, it can vary from protien to protien
  • csimoncsimon Acadiana Icrontian
    edited March 2004
    hm ...I didn't know gromacs had much variation ...maybe we need more experiments like this ...or has there already been a concensus (meaning in a previous thread)?
  • mmonninmmonnin Centreville, VA
    edited March 2004
    Yeah they vary. I dont think we have done much about this. People have asked here and there but not anything big at all.
  • edcentricedcentric near Milwaukee, Wisconsin Icrontian
    edited March 2004
    same unit
    6:22/frame on an XP at 2.069GHz
  • muddocktormuddocktor
    edited March 2004
    I would think that the P4 should fold at about the same PR rating of an AMD proc, such as a P4 2.4 at aprroximately at an XP2400+ speed. One thing that skews any result from me is that both my P4 rigs are HT capable and I run 2 instances, which slows down each instance by a bit.
  • Straight_ManStraight_Man Geeky, in my own way Naples, FL Icrontian
    edited March 2004
    csimon wrote:
    I'm trying to make an observation here ...I'd like to figure out how many mhz/ghz of amd (32bit) processor will fold this protein at appoximately 6:34 - 6:40. At 2.4g the amd must do better that this I'm sure. I'm thinking around the 2g amd range will be equal to the 2.4g p4 in folding but I may be wrong.

    This is what I have so far:

    amd 2000+@1.75 = 7:12
    p4 2.4g = 6:37

    Well, for perspective....

    The P4 @3.1 GHz (at the time it ran, that is the speed the CPU was clocked at) I was running folded it at about 5 min 30 sec per percentage, so complete time was about 9 hours and 10 minutes.

    Guess, to get around 6 min per percentage, or a ten hour complete time, the CPU wanted would be at about the speed of a 2.533 GHz P4 (which has been made available to market). So to get a turnin of less than 10 hours try a 2.66 Notrthwood at OC or at true speed if you get a 2.66 GHz. I have had a 2.66 GHz stable at 3.0 GHz down here in Florida, and in cooler climates I would say 3.1 to 3.2 should be doable for an OC.

    If you want a 6.5 min per percentage average, try a P4 at 2.4 GHz or a 2.2 which should OC fine to 2.66 GHz at least, in the P4 Northwood release range.

    Note that project 563s of several gens and clones run at about 2.6 min per percentage on the box, at 3.2 GHz. 520 is a much slower than normal project for the 500 series, it is by far the slowest I have seen for that series, and project 259 really bogs at almost 10 min per percentage on same box.

    The 800 and 900 series run at less than 4 and greater than 3.4 min per percentage on average, I did not bother mining for details regarding the many I have done, but noted that in skimming through a month of logs to find the most recent one I did on the P4.

    John D-- hoping that helps.
  • csimoncsimon Acadiana Icrontian
    edited March 2004
    I have a 2.533Ghz chip right here but none of my p4's (Dell) are overclockable ...I think mudd hit the nail on the head about the PR rating. I didn't imagine the PR rating was that accurate but apparantly we may find that it is!
  • a2jfreaka2jfreak Houston, TX Member
    edited March 2004
    That's QuantiSpeed rating! ;D;D;D
    (Just Kidding)
    csimon wrote:
    I have a 2.533Ghz chip right here but none of my p4's (Dell) are overclockable ...I think mudd hit the nail on the head about the PR rating. I didn't imagine the PR rating was that accurate but apparantly we may find that it is!
Sign In or Register to comment.